Fix iceberg melt temperature

[cbegeman]

Currently:

• MPAS-Seaice iceberg heat flux term includes both latent heat associated with melting and the heat needed to raise icebergs from -4 degC to 0 degC
• MPAS-Ocean assumes that iceberg mass flux entered at the local freezing point, which is inconsistent with the temperature adjustment above

With this PR:

• MPAS-Seaice iceberg heat flux term includes only the latent heat associated with melting
• MPAS-Ocean assumes that iceberg mass flux entered at the local freezing point, and includes the heat flux needed to raise icebergs from a namelist-derived initial temperature (default -4 degC) to the local freezing point

Local freezing point is the pressure and salinity-dependent freezing point consistent with mushy thermodynamics.

[cbegeman]

Testing

The e3sm_cryo_developer suite has been run on chrys with expected diffs.
The e3sm_prod suite has been run on chrys with diffs only in the SMS_D_Ld1.ne30pg2_r05_IcoswISC30E3r5.CRYO1850.chrysalis_intel case for which config_remove_ais_ice_runoff = .true.

Standalone testing planned for verification but on hold given issues in compass and polaris environments

[cbegeman]

@darincomeau and @xylar You can take a look now or just wait until I post the standalone verification test. Thanks for your input on this!

[darincomeau]

Thanks @cbegeman for putting this together. Agree that prescribed iceberg temperature should be a namelist option.

I'll make a note here that snowcapping iceRunoff is assumed by MPAS-Ocean to be 0 degC coming in (solid), and enters the ocean at 0 degC (liquid). Once we hook up MALI calving fluxes to this field, then we'll have an inconsistency in how icebergs are treated with data icebergs and solid runoff icebergs. Not anything to fix now (changing snowcapping solid ice runoff temperature would be presumably quite disruptive to the larger project), but just something to keep in our minds (especially if we consider down the road having MALI send temperature of calving fluxes).

[cbegeman]

A preliminary look into temperatures at the base of the snowpack suggested that the mean temperature of ice runoff could be significantly lower than -4 degC. I used a single_column test to evaluate the potential impacts of reducing the iceberg temperature from -4 degC to -14 degC and it could result in significant mixed layer deepening at the limit of frazil formation and no frazil ice porosity (along with more frazil formation). So we should keep an eye out for this with future B-case tests.

That result didn't feel right so I found a bug and now here are the differences in the frazil limit:
this branch with iceberg T=-4 degC vs. master: -0.1% frazil formation rate
this branch with iceberg T=-14 degC vs. master: +0.5% frazil formation rate
this branch with iceberg T=-14 degC vs. this branch with iceberg T=-4 degC: +0.7% frazil formation rate

[xylar]

This looks ready for E3SM as far as I'm concerned. Anything else you want to do before we move over?

[darincomeau]

We probably want a bug report issue for this since it's a bug fix PR?

[cbegeman]

This PR addresses E3SM-Project#7110